# facebookresearch/co3d

Tooling for the Common Objects In 3D dataset.

Repository: https://github.com/facebookresearch/co3d
Canonical: https://ross.abutalabs.com/products/co3d
Language: Python
License: NOASSERTION
License Family: other
Last push: 2024-08-14T10:23:58+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1899, "days_push": 749, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1173, forks 87 (observed 2026-08-28T04:03:51.894945+00:00)

## What it is
Tooling for the CO3Dv2 (Common Objects In 3D) dataset, a large-scale collection of real-life multi-view images of everyday objects with camera poses and segmentation masks. It includes scripts for downloading the dataset (5.5 TB full, 8.9 GB single-sequence subset) and supporting the CO3D Challenge evaluation.

## Use cases
- download the CO3Dv2 dataset for 3D reconstruction research
- train single-view or multi-view 3D object reconstruction models
- evaluate models on the Common Objects in 3D Challenge
- get multi-view images with camera poses and segmentation masks
- download a small single-sequence subset for many-view single-sequence experiments

## When to choose
- you need a large real-world multi-view 3D dataset for category-level object reconstruction
- you want to benchmark against published CO3D results or enter the CO3D Challenge
- you need camera poses and segmentation masks for real-life object sequences

## When to avoid
- you lack ~5.5 TB of disk space and only need synthetic 3D data
- you need a ready-made 3D reconstruction model rather than dataset tooling
- you need the original v1 dataset without switching to the v1 branch

## Facets
- artifact type: dataset
- maturity: stable
- function: data-science, machine-learning, computer-vision, image-processing
- domain: computer-vision, machine-learning, deep-learning
- platform: python, windows
- tags: 3d-reconstruction, dataset-tools, benchmark, research-dataset, facebook-research, multiview, datasets, linux, macos

## Member repositories
- facebookresearch/co3d (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:51.894945+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:28:22.627345+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/co3d (fetched 2026-08-28T04:03:51.894945+00:00, sha cf345eec9a53)
- Data as of 2026-08-30T08:39:29.467469+00:00.